JOURNAL OF BACTE1RIOLOGY, Sept., 1965 Vol. 90, No. 3 Copyright @ 1965 American Society for Microbiology Printed in U.S.A,

Physiological Effects of a Constitutive Tryptophanase in Bacillus alvei

J. A. HOCH AND R. D. DEMOSS Department of Microbiology, University of Illinois, Urbana, Illinois Received for publication 8 April 1965

ABSTRACT HOCH, J. A. (University of Illinois, Urbana), AND R. D. DEMoss. Physiological effects of a constitutive tryptophanase in Bacillus alvei. J. Bacteriol. 90:604-610. 1965. -Tryptophanase synthesis in B. alvei is not under the control of tryptophan and is not subject to catabolite repression. Exogenously supplied tryptophan was converted to indole by tryptophanase, and was excreted into the culture medium. The amount of indole excreted was dependent upon the concentration of tryptophan supplied. At intermediate levels of tryptophan (5 to 15 ,jg/ml), the excreted indole was completely reutilized by the cell, in contrast to the result with higher levels. Indole reutil zation was shown to be dependent upon a functional tryptophan synthetase. In the absience of exogenous tryptophan, indole was excreted into the culture medium at an earlier phys- iological age. The early indole was shown not to be a consequence of tryptophanase action. The early indole accompanied uniformly the normal process of tryptophan biosynthesis, and the fission of indole-3-glycerol phosphate was suggested as the origin of the excreted indole.

The enzyme tryptophanase from Escherichia MATERIALS AND METHODS coli has been the subject of considerable research Bacteria. B. alvei ATCC 6348 was obtained from in recent years (Burns and DeMoss, 1962; New- the American Type Culture Collection. A non- ton and Snell, 1964). The enzyme cleaves tryp- mucoid variant which appeared on nutrient agar tophan to indole, pyruvate, and ammonia, with plates was isolated from this strain and used in the pyridoxal phosphate as a cofactor. Tryptophan- experiments reported here. This strain, designated ase in E. coli may serve only in a catabolic role, strain F, has the advantage of being more sus- to metabolizable ceptible to disruption by physical methods than degrading tryptophan pyruvate the parent strain. and excreting indole into the culture medium. Auxotrophic mutants, induced with ultraviolet Newton and Snell (1964) have shown that tryp- light, were selected by the penicillin technique of tophanase may also synthesize tryptophan from Nester, Schafer, and Lederberg (1963). The mu- indole and serine. A mutation to constitutivity tants used in this study are described in Table 1. for tryptophanase in a mutant with the trypto- Accumulated intermediates were identified by phan biosynthetic loci deleted will permit indole paper chromatography. to replace tryptophan as a growth requirement. Culture media. Starter cultures for all experi- Since E. coli is the only organism in which ments were grown on 2% (w/v) Trypticase (BBL), it pH 7.0, supplemented with 10 ,g/ml of thiamine tryptophanase has been studied, was of interest hydrochloride. The mineral salts prepared accord- to determine whether the enzyme was present ing to Smith and Yanofsky (1962) were supple- in other unrelated bacteria. Assuming that pro- mented with 10 ,g/ml of thiamine hydrochloride. duction of indole is a manifestation of tryp- Acid-hydrolysed casein (Nutritional Biochemicals tophanase activity, we have investigated a num- Corp., Cleveland, Ohio), 1% (w/v), was employed as the carbon source. ber of Bacillus species for this characteristic. One Tryptophanase assay. Tryptophanase activity species, B. alvei, produced indole and was found was assayed according to Boezi and DeMoss to exhibit true tryptophanase activity. (1961). The reaction mixture contained: pyri- In the course of investigation of tryptophanase doxal-5-phosphate, 40 pAg; potassium phosphate in B. alvei, its physiological state in the organism buffer (pH 8.0), 300 j,moles; and enzyme solution made to 1.5 ml with distilled water. The reaction was observed to have effects on tryptophan bio- mixture was overlaid with 4 ml of toluene and synthesis; it may fulfill a regulatory role in tryp- shaken at 37 C for 5 min in a 50-ml Erlenmeyer tophan biosynthesis. flask. The reaction was initiated with 0.5 ml of 604 VOL. 90, 1965 CONSTITUTIVE TRYPTOPHANASE IN B. ALVEI 605

0.02 M L-tryptophan and allowed to incubate at TABLE 1. Characteristics of Bacillus alvei mutants* 37 C for 1 hr. The reaction was stopped by adding 0.1 ml of 2 N NaOH and shaking for 5 min. Mutant no. Growth response Accumulation The indole present in the toluene layer was deter- mined according to Yanofsky (1955). Indole for- TS-2 An, In, Try mation in the assay is linear with respect to time TS-15 In, Try IG for at least 5 hr, and is linear with respect to en- TS-19 In,t Try In zyme concentration. One unit of enzyme is defined as that amount of * Abbreviations: An, anthranilic acid; In, enzyme forming 0.1 jAmole of indole per hr. Spe- indole; Try, tryptophan; IG, indole-3-glycerol. cific activity is expressed as units of enzyme per t Leaky on indole. absorbancy unit at 660 m,u of the cell suspension. Determination of indole. To determine indole in TABLE 2. of tryptophan on tryptophanase small samples of the culture, advantage was taken Effect of the high extinction coefficient of the indole-p- synthesis* dimethylaminocinnamaldehyde (PDAC) complex (Scott, 1960). The PDAC reagent consists of 5 Addition Amount Growth 1.activitySpecific parts of a 5% (w/v) solution of PDAC in 95% (v/v) ethyl alcohol and 12 parts of acid-alcohol ,ug/ml (16 ml of concentrated H2SO4 and 200 ml of 95% Tryptophan ...... 0 0.260 3.61 ethyl alcohol). To 0.3 ml of culture, 1.0 ml of 10 0.244 2.79 PDAC reagent was added, and after thorough mix- 50 0.258 3.33 ing the tubes were centrifuged for 5 min at 3,400 X 100 0.250 3.60 g to sediment debris. The clear supernatant fluid 500 0.250 3.28 was removed, and the absorbancy at 625 m, was 1,000 0.201 2.98 read in a Zeiss PMQ II spectrophotometer after Anthranilic acid ...... 200 0.337 3.17 10 min. Indole concentration was estimated from a Indole ...... 50 0.305 3.21 previously constructed standard curve. In some cases the indole was determined after extraction * Cells were harvested after 12 hr, and were into toluene. Either assay gives comparable re- assayed immediately. Growth is presented as sults. absorbancy at 660 m,u; specific activity is ex- Determination of tryptophan. Tryptophan was pressed as units of enzyme per absorbancy unit determined in culture supernatant fluids by the at 660 m,u. method of Frank and DeMoss (1957) with par- tially purified E. coli tryptophanase. Measurement of absorbancy. The absorbancy of (Table 2). Indole and anthranilic acid similarly cultures was measured at 660 m,u, either in a Zeiss have no effect. Such a situation can be envisioned PMQ II spectrophotometer or in an Evelyn photo- if tryptophan is overproduced by the organism, electric colorimeter. thus rendering the exogenous level of tryptophan Measurement of radioactivity. Indole-2-C14 insignificant. On the contrary, tryptophan auxo- (Calbiochem) was converted enzymatically to L-tryptophan-2-C14. In a typical experiment, a trophs grown on a limiting concentration of tryp- sample of culture was sedimented at 8,700 X g for tophan (e.g., 10 ,ug/ml) do not have enzyme levels 15 min. The cells were suspended in 5% (w/v) lower than the wild type. An alternative expla- trichloroacetic acid, and were allowed to stand at nation would suggest that the organism lacks a 4 C for at least 30 min. A sample of this cell sus- tryptophan permease. Although no permease pension was filtered on a membrane filter (no. studies were undertaken, this possibility is not B-6, Carl Schleicher & Schuell Co., Keene, N.H.), likely because copious amounts of indole accu- and the filter was allowed to dry. The supernatant mulate in the presence of exogenous tryptophan. fluid from the initial centrifugation was extracted with toluene, and samples of the toluene layer Thus, tryptophanase synthesis is not under were assayed for indole and radioactivity. A Tri- the control of tryptophan but rather is constitu- carb Liquid Scintillation Spectrometer was em- tive. It might be noted that the specific activity ployed for all radioactivity counting. of fully induced E. coli is approximately 200-fold higher than that of B. alvei. RESULTS Effect of tryptophan on indole excretion. When B. alvei was grown in the presence of Effect of tryptophan on enzyme level. E. coli tryp- tryptophan, tophanase is inducible by high levels of trypto- indole was excreted into the culture medium and phan; in its absence, only a minute amount of presumably arose from tryptophanase action. enzyme is synthesized. Conversely, in B. alvei Indole also appeared in the absence of exogenous the level of exogenous tryptophan has no influ- tryptophan. Figure 1 shows the indole-excretion ence on the specific activity of tryptophanase kinetics for various levels of tryptophan. When 606 HOCH AND DEMOSS J. BACTERIOL. tions (20 ,ug/ml) increased the absolute amount of indole formed but did not appreciably alter the kinetics of indole excretion. At lower concen- trations, anthranilate seemed to decrease the amount of indole formed. This observation lends support to a role for tryptophanase in early indole formation, since Burns (Ph.D. Thesis, Univ. of w 0 Illinois, Urbana) has shown that anthranilate is a competitive inhibitor of the E. coli trypto- 0 phanase. However, Gibson and Yanofsky (1960) showed that indole-3-glycerol phosphate (IGP) E synthetase is also inhibited by anthranilate. E Hence, the reaction is more probably a function 4 of a decreased IGP level. Indole excretion in tryptophan auxotrophs. Fig- 4 ure 2 shows the indole excretion kinetics of a mutant (TS-2) blocked before anthranilic acid. 2 4 6 8 l0 When this mutant was grown on 10 ,g/ml of TIME, HRS tryptophan, the indole-excretion kinetics were FIG. 1. Effect of exogenous tryptol han on indole- similar to the wild type at the same concentra- excretion kinetics during growth. Indole concen- tion of tryptophan. When anthranilic acid was tration is expressed as miAmoles pier milliliter of the supplement, the mutant showed indole-ex- supernatant fluid after sedimentati(on of the cells. cretion kinetics corresponding to the wild type Tryptophan (per milliliter): none, 0; 1 jg, 0; when the latter was grown in the absence of 56Ag, E;10 Ag, +. tryptophan. Thus, it may be concluded tenta- tively that the appearance of early indole is a no tryptophan was present, indo]le appeared at function of the biosynthetic pathway and accom- the onset of exponential growth, rose to a maxi- panies uniformly the normal process of trypto- mum, and was reutilized completely before max- phan biosynthesis. This hypothesis suggests that imal growth was attained. Trypto)phan at levels the early indole arises from the fission of IGP. of 5 Mg/ml or greater effects a comiplete inhibition In Fig. 3, the indole-exeretion kinetics for of early indole excretion. Under these conditions, tryptophan synthetase mutants are shown. Mu- indole excretion was delayed, and, in the cultures tant TS-15 corresponds to an A protein mutant which contained intermediate lev7els of trypto- in E. coli (i.e., IGP is not converted to indole; phan (5 or 10 ,Ag/ml), subsequer itly was reuti- Yanofsky, 1960), and it is evident that this mu- lized. In the presence of either high (500 sg/ml) tant shows kinetics similar to those of the wild or low (5 ,ug/ml) levels of exogenous tryptophan, type in the presence of tryptophan. Thus, we con- indole excretion occurs at the samie culture age. clude that the late indole formation is solely a The indole excretion observed mLay derive from function of tryptophanase action. Mutant TS-19 a catabolic or an anabolic origin. IIn the presence corresponds to a B protein mutant in E. coli (i.e., of exogenous tryptophan, degradabtion catalyzed indole is not converted to tryptophan) in both by tryptophanase could account for the indole growth response and accumulation. In this case, formed. In the absence of tryptop]han, the indole the late indole that appears is not reutilized. might arise from indole-3-glycerol phosphate via Thus, we conclude that the late indole is con- the biosynthetic pathway. Indeed at low levels verted to tryptophan upon reutilization and is of tryptophan both mechanisms rriight be opera- not converted to IGP or otherwise metabolized. tive. These observations raise an additional Fate of isotopically labeled L-tryptophan. To question concerning the fate of the reutilized demonstrate unequivocally that late indole was indole. Clearly, tryptophan is not the only pos- derived from exogenous tryptophan, a mutant sible product of indole reutilizatio:n. In attempts (TS-2) blocked before anthranilate was grown to answer these questions, recours;e was made to with L-tryptophan-2-C'4. Since the 2 position of tryptophan auxotrophs and to exj periments with isotopically labeled tryptophan. the indole ring is lost in an anthranilate cycle, Effect of anthranilic acid and ixndole on indole the experiment also allowed us to rule out the excretion. The presence of indole in the growth possible existence in B. alvei of such a cycle medium did not prevent the appiearance of the (Matchett and DeMoss, 1963). Figure 4 shows early indole, but the amount whiclh appeared was the indole-excretion kinetics in this experiment. decreased. Anthranilic acid at high concentra- Clearly, the indole arose from the labeled tryp- VOL. 90, 1965 CONSTITUTIVE TRYPTOPHANASE IN B. ALVEI 607

TIME, HOURS

- 1.5 0 i x w

0 0 0 z 0 1.0 2 -J0 E 0 E 0.5k_

-0 0 1 2 3 4 5 6 7 8 9 TIME, HOURS FIG. 2. Indole-exeretion kinetics for mutant TS-2. (A) Indole excretion with 10,ug/ml of anthranilic acid as the supplement; (B) indole excretion with 12 ,ug/ml of L-tryptophan as the supplement. Growth, 0; in- dole, 0; exogenous tryptophan, +. tophan supplied exogenously, and the specific synthetic tryptophan in the metabolic pool, and activity of the excreted indole (1,402 counts per label would appear in the indole. On the other min per m,umole) was the same as the trypto- hand, if early indole were a consequence of IGP phan-indole (1,373 counts per min per m,umole). fission only, the excreted indole would be unla- The latter specific activity was taken from the beled. The results of this experiment are depicted 1-hr tryptophan value, since a small quantity of in Fig. 5. The indole that appeared in this ex- unlabeled tryptophan was brought in with the periment was not labeled. The tryptophan added inoculum. was of an isotopic specific activity such that 1% Origin of early indole. The addition of 1 ,ug/ml could be detected easily in the indole. We con- of tryptophan to the culture neither repressed clude tentatively that early indole is not a func- the formation of early indole nor altered the tion of tryptophanase action, but rather is de- kinetics of indole excretion (Fig. 1). If the indole rived from IGP fission. is a consequence of tryptophanase action on bio- synthetically produced tryptophan, the addition DISCUSSION of 1 ,ug/ml of labeled tryptophan should allow The physiological status of tryptophanase in detection of label in the indole, providing the B. alvei differs from the same enzyme in E. coli. internal metabolic pool of tryptophan is a finite The enzyme in B. alvei appears to be fully con- size. That is, labeled exogenous tryptophan stitutive and not affected by exogenous trypto- would be expected to mix with endogenous bio- phan levels. In addition, "catabolite repression" 608 HOCH AND DEMOSS J. BACTERIOL.

0 2 4 6 8 10 TIME, HOURS FIG. 3. Indole-excretion kinetics for tryptophan synthetase mutants. Growth, @; indole for mutant TS-15, 0); indole for mutant TS-19, +.

(Magasanik, 1961) is not operative on this en- zyme system (Hoch, Ph.D. Thesis, Univ. of Illi- I nois, Urbana). z 2.0 w 4 -J The possibility that tryptophanase in B. alvei IL a 12 0. 2 serves in lieu of a normal tryptophan synthetase z I~- C1 for the synthesis of tryptophan (Newton and 0 I.- 40- 1.0- Snell, 1965) was eliminated. A mutant (TS-19) E unable to use indole for growth continued to ex- E A crete indole in the presence of exogenous trypto- O0 phan. 2 44 6 8 10 The constitutive production of tryptophanase TIME, HRS in B. alvei is manifested in its effect on trypto- FIG. 4. Indole-excretion kinetics for mutant TS-2 phan utilization. The addition of tryptophan, growing on L-tryptophan-2-C14. Exogenous trypto- even in low concentrations, results in a subse- phan, 03; indole, *; toluene-extractable radio- quent excretion of indole into the culture me- activity per milliliter of supernatant fluid, 0. dium. The amount of indole excreted depends on the original tryptophan concentration. The the early phase of growth when presumably concentration of excreted indole represents at enough tryptophan was present in the intra- most 15 to 20% of the original tryptophan sup- cellular amino acid pools from the starter culture, plied, when intermediate levels (5 or 10 ,ug/ml) approximately 100% of the tryptophan taken up were used. Thus, not all of the tryptophan taken was excreted as indole. When the exogenous up by the cell is degraded to indole. This effect tryptophan was exhausted, the extracellular in- is shown in Fig. 4. The extracellular indole con- dole concentration was greatest and at this time centration increased at the rate of approximately the indole was rapidly reutilized. 2 m,umoles/30 min while the tryptophan concen- The reutilization of indole has been shown to tration decreased at the rate of approximately be a consequence of tryptophan synthetase activ- 20 m,moles/30 min. In this case, 10% of the ity. A mutant (TS-19) unable to grow on indole tryptophan taken up was excreted as indole. In did not reutilize the indole which was produced. VOL. 90, 1965 CONSTITUTIVE TRYPTOPHANASE IN B. ALVEI 609

6

0 x E 0 co -9 a.IL

to Ct)0 J 0

0 2 4 6 8 10 TIME, HOURS FIG. 5. Indole-excretion kinetics in wild-type Bacillus alvei with 1 jig/mi of L-tryptophan-2-C14 supple- ment. Growth, *;indole, 0 ; trichloroacetic acid-insoluble radioactivity per milliliter of culture, +; toluene- extractable radioactivity per milliliter of supernatant fluid, X.

Thus, indole is not metabolized to products other ACKNOWLEDGMENTS than tryptophan. This investigation was supported by Public It seems reasonable to draw a general analogy Health Service research grant E-2971 from the between tryptophanase systems and the anthra- National Institute of Allergy and Infectious Dis- nilate cycle of Neurospora. Matchett and DeMoss eases. The senior author was supported by a traineeship from a General Microbiology Train- (1964) showed the presence of two amino acid ing Grant (GM-510) from the National Institute pools in this organism, the metabolic pool and of General Medical Sciences. the expandable pool. Tryptophan in the latter pool is available for degradation via the anthra- LITERATURE CITED nilate cycle, whereas the smaller metabolic pool BAUERLE, R. H., M. FREUNDLICH, F. C. STORMER, AND H. E. UMBARGER. 1964. Control of isoleu- is assumed to mediate regulation of the biosyn- cine, valine and leucine biosynthesis. II. End- thetic enzymes. It is conceivably advantageous product inhibition by valine of acetohydroxy for the B. alvei cell to regulate quickly the size of acid synthetase in Salmonella typhimurium. a tryptophan pool by means of tryptophanase. Biochim. Biophys. Acta 92:142-149. BOEZI, J. A., AND R. D. DEMoss. 1961. Properties If tryptophan proved inhibitory to another of a tryptophan transport system in Escherichia process in the cell, it would be especially advan- coli. Biochim. Biophys. Acta 49:471-484. tageous to maintain sensitive control of the tryp- BURNS, R. O., AND R. D. DEMoss. 1962. Proper- tophan concentration. Evidence for this possibil- ties of tryptophanase from Escherichia coli. ity can be adduced from the valine inhibition of Biochim. Biophys. Acta 65:233-244. FRANK, L. H., AND R. D. DEMOSS. 1957. Specific isoleucine biosynthesis in Salmonella typhimur- enzymic method for the estimation of L-trypto- ium, reported by Bauerle et al. (1964). The next phan. Arch. Biochem. Biophys. 67:387-397. best inhibitor of the valine inhibitable enzyme GIBSON, F., AND C. YANOFSKY. 1960. The partial (acetohydroxy acid synthetase) is tryptophan. purification and properties of indole-3-glycerol case it phosphate synthetase from Escherichia coli. Clearly, in this would be advantageous to Biochim. Biophys. Acta 43:489-500. keep tryptophan at or below a particular low MAGASANIK, B. 1961. Catabolite repression. Cold level within the cell. Spring Harbor Symp. Quant. Biol. 26:249-256. 610 HOCH AND DEMOSS J. BACTERIOL.

MATCHETT, W. H., AND J. A. DEMOSS. 1963. and tryptophan synthetases in Escherichia coli. Direct evidence for a tryptophan-anthranilic J. Bacteriol. 89:355-364. acid cycle in Neurospora. Biochim. Biophys. SCOTT, T. A. 1960. An enzymic method of estimat- Acta 71:632-642. ing L-tryptophan and the interference by a 3- MATCHETT, W. H., AND J. A. DEMoss. 1964. indolinylidene-3-indolyl-4-pyridylmethane. Bio- Physiological channeling of tryptophan in Neu- chem. J. 75:7P. rospora crassa. Biochim. Biophys. Acta 86:91-99. SMITH, 0. H., AND C. YANOFSKY. 1962. Enzymes NESTER, E. W., M. SCHAFER, AND J. LEDERBERG. involved in the biosynthesis of tryptophan, 1963. Gene linkage in DNA transfer: a cluster of p. 794-806. In S. P. Colowick and N. 0. Kaplan genes concerned with aromatic biosynthesis in [ed.], Methods in enzymology, vol. 5. Academic Bacillus subtilis. Genetics 48:529-551. Press, Inc., New York. NEWTON, W. A., AND E. E. SNELL. 1964. Catalytic YANOFSKY, C. 1955. Tryptophan synthetase from properties of tryptophanase, a multifunctional Neurospora, p. 233--238. In S. P. Colowick and pyridoxal phosphate enzyme. Proc. Natl. Acad. N. 0. Kaplan [ed.], Methods in enzymology, Sci. U.S. 51:382-389. vol. 2. Academic Press, Inc., New York. NEWTON, W. A., AND E. E. SNELL. 1965. Forma- YANOFSKY, C. 1960. The tryptophan synthetase tion and interrelationships of tryptophanase system. Bacteriol. Rev. 24:221-245.